Study of Refining Effect of Mixed Pulps Using Refiner Plates with Different Bar Patterns

Abstract

A new lightweight vertical bar plate was developed by inserting individual bars into the plate base rather than the typical sand casting method. The lightweight vertical bar plate has a very sharp bar edge because vertical bars are inserted instead of trapezoidal bars. The refining effects of the mixed stock with softwood bleached kraft pulp (Sw-BKP) and hardwood bleached kraft pulp (Hw-BKP) were compared using two types of lightweight vertical bar plates with cutting edge lengths of 55 km/s (PA55) and 97 km/s (PB97). The fine bar plate (PB97) with large CEL (PB97) was very effective in reducing the refining energy required to reach the final freeness regardless of the mixing ratios of Sw-BKP and Hw-BKP compared to the plate with a small CEL (PA55). PB97 also consumed less refining energy than PA55 until proper sheet strength was achieved. In particular, irrespective of the mixing ratios of Sw-BKP and Hw-BKP, the tensile strength of sheets at the final freeness was greater in PB97 than in PA55. For stock throughput during refining, PA55 with the wide groove was almost similar to PB97 with the narrow groove.

1. Introduction

A refining system should be selected with consideration of the properties of the end products, available pulp types, and production capacity range. Refining during stock preparation is commonly performed separately for different types of pulps because they show varying refining behaviors. However, other types of pulp are sometimes mixed and processed before refining. In either case, the number of refiners depends on the target freeness and production capacity [1]. For example, in a toilet tissue mill, only one refiner is required because the pulp fibers are lightly refined, but in the case of glassine paper, 5–6 refiners are often needed for extensive refining [2].

Currently, both independent refining systems and mixed refining systems are widely used in paper mills. New large paper machines use independent refining, and older small paper machines use mixed refining. As is well known, independent refining and mixed refining each have their own advantages. In some pulp blends, independent refining develops better strength while consuming less refining energy than mixed refining [1,3].

However, for some pulp blends, mixed refining may give better results. In addition, mixed refining may be more suitable for some pulp blends, so the dimension and bonding ability of the fibers may determine whether independent refining or mixed refining is appropriate [2,4]. If softwood fibers with long fiber lengths and thin fiber walls are mixed with hardwood fibers with short fiber lengths and thin fiber walls, the hardwood fibers are loosely bound to the softwood fibers [5]. As hardwood pulp fibers pass through the refiner grooves without enough modification, independent refining becomes more desirable in these pulps [1,6]. Conversely, if short, thick softwood pulp fibers are mixed with long, thin hardwood pulp fibers, the hardwood pulp fibers are firmly bound to the softwood fibers and are protected [5]. In this case, if mixed refining is applied, good refining results can be expected due to inter-fiber contact.

For this reason, a combined system of mixed refining after separate refining of different pulps provides good alternatives because the benefits from both independent and mixed refining systems can be obtained.

In the refining process, the pulp fibers undergo compression and shear forces to change their properties, which will vary depending on the initial fiber properties or the specifications of the refiner [7,8]. Therefore, refining should be applied differently depending on the properties of the fibers. In general, long fibers can be shortened through high-intensity refining (free beating) to achieve better formation, and excessive fiber shortening in short fibers can be minimized by low-intensity refining (wet beating) to preserve fiber length [2,9,10,11,12]. Currently, most paper manufacturers prefer low-intensity refining as the use of hardwood pulp and recycled pulp increases [13,14].

The most frequently used way to evaluate refining intensity is the Specific Edge Load (SEL). Plate bar fillings are the most critical factor in determining refining intensity suitable for papermaking [15]. According to the SEL equation, a straightforward way to achieve low-intensity refining is to increase the number of bars in the refiner plate, increasing the Cutting Edge Length (CEL) [16]. However, even if the CEL is the same, the refining result may vary depending on the bar filling or bar design. In general, refiner plates have been manufactured using conventional casting technology, which requires a 3°–5° minimum draft angle for plate bars [17,18]. This small angle helps with the removal of the pattern without damaging the sand mold. However, with the recent introduction of technologies capable of manufacturing vertical bars without the draft angle, it has become possible to use refiner plates that are advantageous in terms of bar sharpness and stock throughput. Gu et al. developed a lightweight bar plate without a draft angle using different metal alloys [19].

In this study, when refining the mixed pulps, the difference in the development of refining effects was investigated between a plate with a vertical bar shape and a plate with a trapezoidal bar shape. If it is confirmed that the refining effects of the lightweight vertical bar plate on the mixed pulps are excellent, this study is expected to be used as a fundamental work to introduce a future refining system for paper quality improvement and energy saving in modern paper mills.

2. Experimental

2.1. Raw Materials

Moorim Paper Inc. mill in Jinju, Korea supplied the Hardwood bleached kraft pulp (Hw-BKP) and softwood bleached kraft pulp (Sw-BKP) for stock preparation. The characteristics of Hw-BKP and Sw-BKP are shown in

Table 1. Fiber properties of raw materials.

	Length-Weighted Mean Fiber Length (mm)	Length-Weighted Fine Contents (%)	Fiber Width (µm)	Fiber Coarseness (mg/m)	Freeness (mL CSF)
Hw-BKP	0.7	9.5	20.6	0.46	640
Sw-BKP	2.2	5.1	34.0	0.18	790

. These dried pulps, torn into pieces about 30 mm square, were soaked in distilled water for at least 4 h before disintegration. The Hw-BKP and Sw-BKP were blended in 3:7, 5:5, and 7:3 ratios and then refined.

2.2. Manufacturing a Novel Refiner Plate without a Draft Angle

Refiner plates manufactured using conventional sand casting usually require 2 to 5 degrees of draft angle to prevent damage to the sand mold during the pattern removal step (see Figure 1). If the draft angle can be eliminated, the new plate will have several advantages over trapezoidal bar plates in terms of stock throughput, energy saving, plate life, and bar edge sharpness [20,21].

However, it was not possible to remove the draft angle with the general sand casting method, so a new manufacturing method had to be developed. A stainless alloy plate with excellent wear and corrosion resistance was cut with a laser to about 15 mm high and the desired bar length to manufacture a plate without a draft angle. Each cut bar was heat-treated at about 1500 °C and then inserted into the pre-made grooves of the metal mold (refer to Figure 2a). The flask was placed in the metal mold with the inserted bars and filled with resin-coated sand. The flask was hardened at 200–500 °C to make a sand mold. The flask for the base plate was also prepared in a similar way. When the cooled sand-mold flask was lifted, the bars inserted in the metal mold were transferred to the flask and combined with the base-plate flask.

After the molten aluminum alloy was placed in the inlet between the flasks and cooled (refer to Figure 2b), the sand mold adhering to the finished plate was removed, followed by post-treatment to adjust the plate base surface and the bar height uniformly.

Since the refiner plate is manufactured with laser-cut stainless-steel bars inserted into the plate base, the bar has a rectangular shape in a lateral view without the draft angle, unlike a conventional casting plate (refer to Figure 3).

Figure 4 shows the two types of novel refiner plates without the draft angle used for this study. The refining effects of the mixed pulps were investigated using these two novel plates with different bar fillings. CELs for the two plates were calculated according to the guidance of ISO/TR 11371 as follows:

$$\mathit{SEL}(\mathrm{J}/\mathrm{m})= \frac{{\mathit{P}}_{\mathit{t}}{-\mathit{P}}_0}{{\mathit{Z}}_{\mathit{r}}{\times \mathit{Z}}_{\mathit{st}}\times \mathit{l}\times \mathit{n}}= \frac{{\mathit{P}}_{\mathit{t}}-{\mathit{P}}_0}{\mathit{CLF}\times \mathit{n}}= \frac{{\mathit{P}}_{\mathit{t}}-{\mathit{P}}_0}{\mathit{CEL}}$$

(1)

where SEL is the Specific Edge Load (SEL, J/m, W·s/m), $P_t$ is the total load power (kW), $P_0$ is the no-load power (kW), and n is the rotational speed (rev/s). $Z_r$ and $Z_{st}$ are the total numbers of rotor and stator bars, l is the bar length (km), CLF is the cutting length factor (km/rev), and CEL is the cutting edge length (km/s).

Table 2. Comparison of the refiner plates with different bar fillings.

	Plate A (PA55)	Plate B (PB97)
Plate
Bar dimension (mm)
Bar angle (°)	20	40
Bar number/segment	67	160
Segment No.	3	3
CEL (km/s)	55	97
Weight/segment (g)	924.9	794.5

compares the dimensions and CEL of the two plates with different bar fillings. Plate A (PA55) had 67 bars, significantly fewer than Plate B (PB97), with 160 bars, producing a much smaller CEL than PB97. CEL has been regarded as the predominant factor in determining SEL. Low-intensity refining becomes possible with the increased CEL of the refiner plate.

2.3. Refining

Before refining, the pulps were disintegrated using the Valley beater with no load according to ISO 5264-1. After disintegration, the consistency of the disintegrated stock was adjusted to about 2–4% for refining. Refining was consecutively conducted at 1500 rpm up to around 180 mL Canadian Standard Freeness according to the ISO/TR 11371 using a laboratory single-disc refiner (KOS1 Co., Gimhae, Korea). The refiner was fitted with two different plates, such as PA55 and PB97 (refer to Figure 5).

Specific Refining Energy (SRE) consumed during refining was measured by Equation (2) as follows:

$$\mathrm{Specific}\mathrm{Refining}\mathrm{Energy} \left(\frac{\mathrm{kWh}}{\mathrm{t}}\right)=\frac{P_{tot}-P_0}{f \times \mathrm{c}}$$

(2)

where $P_{tot}$ is the total refiner load power (kW), $P_0$ is the refiner no-load power (kW), $f$ is the flow rate (L/h), and $\mathrm{c}$ is the stock consistency (t/L, based on oven-dried pulp weight).

2.4. Analysis of Pulp and Paper Properties

The water retention value (WRV) of pulp fibers was measured based on ISO 23714. A fiber quality analyzer (FQA-360, Optest Equipment Inc., Hawkesbury, ON, Canada) was used to measure the mean fiber length and fines content of pulp fibers.

Handsheets with a basis weight of 70 g/m² were prepared using a laboratory handsheet machine according to ISO 5270. The physical properties of the handsheets, including caliper, tensile, burst and tear strength were measured according to ISO 534, ISO 1924-2, ISO 2758, and ISO 1974.

3. Results and Discussion

3.1. Fiber Properties

Refining is easily assessed using the drainage rate of the pulp fibers, which is closely related to the external and internal fibrillation of the pulp fibers, delamination of the fiber wall, and shortening of the fibers [22]. A high and fast drainage rate means a high freeness. Figure 6 compares the refining energy required to achieve the target freeness (approximately 180 mL CSF) when the mixed pulps were individually refined by the two plates. As the mixing ratio of hardwood pulp fibers increased, the refining energy required to reach the target freeness decreased irrespective of the bar fillings of the plates. In the case of Sw-BKP, which was composed of long fibers, the most refining energy was consumed until the target freeness was reached irrespective of plate filling, but in the case of PB97, the refining energy was reduced by about 50% compared to PA55.

When the refining energy consumed up to the target freeness was compared, it was confirmed that the energy consumption of PA55 was significantly reduced, as Hw-BKP was mixed. It is worth noting that the refining energy was remarkably decreased when Hw-BKP was mixed with Sw-BKP compared to Hw-BKP alone. PA55 has a wide groove between the bars, so short hardwood fibers are more likely to pass unrefined. However, as hardwood fibers were mixed with long softwood fibers, the time for the mixed fibers to pass through the grooves was delayed, so it was considered that refining was appropriately performed (refer to Figure 7a). As the long softwood fibers held the short fibers to form large flocs, these flocs were more easily caught on the bar edges of the plate and appeared to be refined quickly. In the end, it was confirmed that, as the CEL of the refiner plate became smaller, it was more advantageous to refine the mixed pulps with long and short pulp fibers than to refine a single pulp.

However, PB97 showed different refining behavior from PA55. As the mixing ratio of Hw-BKP increased, the refining energy required up to the target freeness decreased sequentially. In PB97, with narrower grooves, it was considered that more softwood fibers were mixed to form larger or harder flocs, which took more time to pass through the plate grooves, and consequently, more refining energy was consumed [6]. On the other hand, the soft flocs with short fibers passed through the plate grooves more easily and caught on the bar edges so that they were refined quickly.

Figure 8 shows the CEL effect of the refiner plates on the mean fiber length and fines content of the mixed pulps refined up to about 180 mL CSF. It is well known that high-intensity refining tends to induce fiber shortening [23]. When only Sw-BKP was refined, PA55 with smaller CEL generated more fiber cutting than PB97 with higher CEL, leading to a shorter mean fiber length. However, there was no significant difference in mean fiber length and fines content between PA55 and PB97 even though the hardwood fibers were mixed.

Figure 9 is a graph comparing the WRV of the mixed pulps refined up to about 180 mL CSF. When only Sw-BKP and Hw-BKP were refined independently, there was no remarkable WRV difference in both PA55 and PB97. However, as Sw-BKP and Hw-BKP were mixed, the WRVs of the mixed pulps were greater than those of the single pulp, regardless of CEL. As Sw-BKP was mixed with Hw-BKP, long fibers formed flocs with short fibers during refining, and the flocs absorbed the physical impacts from the bar edges [1,6]. As a result, it was considered that the fiber shortening during refining seemed to be reduced, and the bonding ability of refined pulp fibers with water molecules was increased with internal and external fibrillation [24]. It was greatly interesting to note that PB97 had WRVs similar to PA55 despite lower refining energy consumption, regardless of the mixing ratios of Sw-BKP and Hw-BKP.

In conclusion, irrespective of the plate CEL, when refining using a plate with sharp bar edges, fiber properties of the mixed pulps could be developed faster while reducing refining energy consumption than refining Sw-BKP and Hw-BKP separately. In addition, it was confirmed that the plate with greater CEL contributed to the faster and better development of fiber properties of the mixed pulps.

3.2. Physical Properties of Paper

It is necessary to refine or beat pulp fibers to develop appropriate fiber properties for papermaking. After refining, pulp fibers are flattened and become more flexible, and have a higher area available for bonding [25]. Figure 10 shows the effect of CEL on the bulk of the sheets made from the mixed pulps. In general, the paper bulk was negatively affected by refining irrespective of pulp mixing ratios or plate CEL. As Sw-BKP was mixed with Hw-BKP, more refining energy was consumed to achieve the same bulk. In particular, PB97, with larger CEL, consumed about half of the refining energy compared to PA55 to reach the same bulk at each mixing ratio.

It was confirmed that, when refining the mixed pulp with a fine bar plate with a very large CEL, paper with good bulk could be produced while consuming less refining energy.

Figure 11 compares the effect of CEL on the tensile strength of the sheets made from mixed pulps. Irrespective of the CEL difference, the refining effects for Sw-BKP were slowly developed, but it had the greatest tensile strength at the target freeness. On the other hand, the refining effects for Hw-BKP were rapidly developed, but the tensile strength at the target freeness was the lowest.

In the case of PA55, the refining energy for the mixed pulps of Hw-BKP and Sw-BKP required up to the same tensile strength was less than the energy for the independent refining of Sw-BKP and Hw-BKP. PB97 also showed a similar trend, but the refining energy required for the same tensile strength was less than that of PA55. The tensile strength in the target freeness was greater in PB97 than in PA55, irrespective of the mixing ratios of Sw-BKP and Hw-BKP. As more Sw-BKP was mixed, greater refining energy in PB97 was required at the same tensile strength [3]. It was finally confirmed that the fine bar plate with a very large CEL was highly advantageous in strength development and energy saving.

Figure 12 compares the effect of CEL on the tear strength of paper prepared from the mixed pulps of Sw-BKP and Hw-BKP. It is known that tear resistance is closely related to the length of the pulp fibers. Longer fibers prepare weakly bonded sheets, but the tear strength of these papers is improved. Refining contributes to the production of well-bonded sheets, but improved bonding strength reduces the tearing resistance [26]. PA55 and PB97 plates with sharp bar edges improved tear strength irrespective of the pulp mixing ratios during refining. However, PA55 had better tear resistance as the mixing percentage of Sw-BKP increased. In particular, when Sw-BKP and Hw-BKP were mixed at a ratio of 7:3, it showed much better tear resistance than Sw-BKP under less refining energy. These mixed pulps were generally similar to the tear resistance behavior of typical softwood pulp. The mixing of short fibers with long fibers seemed to help improve interfiber bonding [27].

Unlike PA55, when refining with PB97, tear resistance increased irrespective of the pulp mixing ratios of Sw-BKP and Hw-BKP, and the sheets prepared from the pulps with more Sw-BKP showed better tear resistance than those prepared from Hw-BKP alone at each refining step. It was interesting to note that the sheet with a mixing ratio of Sw-BKP and Hw-BKP of 7:3 also displayed the best tear resistance. In particular, much less refining energy in the final tear strength was much less consumed in the fine bar plate PB97 than in PA55.

3.3. Stock Throughput

Figure 13 compares the stock throughput of the mixed pulps passing through the plate grooves of PA55 and PB97 during refining. It was found that it was much easier for pulp fibers to pass through the refiner plates with rectangular grooves between the bars than the plates with trapezoidal grooves [18]. However, as shown in Figure 13, different stock throughputs were obtained depending on the CEL values between the plates having the same rectangular grooves. Although PA55 with a smaller CEL has slightly wider grooves than PB97, when compared to the same refining energy, the stock throughput of PA55 was smaller or similar to that of PB97. As a result, it was confirmed that the plates with a vertical bar shape were not substantially different in terms of stock throughput despite different CEL values.

4. Conclusions

The refining effects of the mixed pulps were compared by manufacturing two types of lightweight plates having vertical bar shapes with different CELs. The fine bar plate with very large CEL (PB97) was very effective in reducing the refining energy required to reach the final freeness irrespective of the mixing ratios of Sw-BKP and Hw-BKP compared to the plate with a small CEL (PA55). In addition, PB97 required less refining energy than PA55 until proper sheet strength was obtained. In particular, irrespective of the mixing ratios of Sw-BKP and Hw-BKP, the tensile strength of sheets at the target freeness was greater in PB97 than in PA55. For stock throughput during refining, PA55, with the wide groove, was almost the same as PB97, with the narrow groove. In conclusion, it was confirmed that the vertical bar plate with sharp bar edges with larger CEL was more effective in the development of the refining effects, as well as saving energy on refining, for the mixed pulps of Sw-BKP and Hw-BKP.